FlashMind — Smart Flashcard Generator

## Inspiration
Studying is hard. Re-reading notes is boring. I wanted a tool that could instantly turn any text into something actually useful — structured,
quizzable flashcards. The idea was simple: paste your notes, let AI do
the heavy lifting.

## What I Learned

• How to integrate Groq's Llama 3.3 70B for fast AI generation
  
• Implementing the SM-2 Spaced Repetition algorithm, where the next
  review interval is calculated as:
  

$$I(n) = I(n-1) \times EF$$

where $EF$ (ease factor) adjusts based on how well you recalled the card.                                                      

• Building a full-stack app with React + Node.js + Express
  
• Designing a clean, accessible UI with dark mode support

## How I Built It

• Frontend: React 18 + Vite + Tailwind CSS
  
• Backend: Node.js + Express
  
• AI: Groq API (Llama 3.3 70B) for flashcard generation
  
• Speech: Initially integrated ElevenLabs TTS API for realistic voice
  output, but discovered the free tier doesn't support premade voices via API.
  Switched to the browser's built-in Web Speech API — free, offline,
  and works on all modern browsers.
  
• Algorithm: SM-2 spaced repetition for smart review scheduling
  

## Challenges

• Getting the AI to return consistent JSON without extra formatting
  
• Switching between multiple AI providers (tried Grok/xAI, settled on Groq)
  
• Building a multi-deck system with per-deck progress tracking
  
• Making the UI feel polished and Play Store-ready on a tight timeline

Built With

• express.js
• groq
• node.js
• react
• tailwindcss
• vite